Face and iris recognition apps both thrilling and threatening

TORONTO – There will soon come a time, not so far into the future, when smartphone users will be able to hold up their device before a crowd of people and instantly find friends and acquaintances with a single click or swipe across a touch screen.  

While it sounds rather magical, facial and iris recognition applications are very real. Some advanced identification apps are already available on social media and in the mobile marketplace, while many others are well into later stages of development.

Facial recognition apps act like detectives, scanning a crowd to identify people based on information stored in a pre-existing database of photographs and videos.

It’s the kind of technology that would have helped Vancouver police quickly identify rioters and anarchists hell-bent on destroying parts of the city after the Canuck’s Stanley Cup loss in June.

Similar technology is already in use in about 40 law enforcement units across 14 U.S. states. Police officers can use a specially-designed iPhone to snap a photo or scan a suspect’s iris, then literally tap into a database to access that person’s information, including criminal records. The Mobile Offender Recognition and Information System (MORIS) is being called a “game changer” by officers already using the technology, which costs about US$3,000 per unit.

“Iris scanning technology has become incredibly sophisticated in the past few years,” explains Amy Webb, CEO of Webbmedia Group, an international digital strategy consulting firm based in Washington, DC.

“In a sense, it’s not that different from taking fingerprints, except that MORIS attaches to an iPhone and does real-time scanning on the spot,” she explains.

Come 2014, police officers in Rio de Janeiro will be outfitted with special helmets containing iris scanners to identify and track would-be soccer hooligans attending World Cup matches.

More than a friend-finder 

Thanks to an app by Face.com, facial recognition technology is currently in use on Facebook under the guise ‘Photo Finder,’ a built-in program that helps users detect, recognize and tag photos of themselves and friends across their entire social network.

Google is capitalizing on this technology with the recent purchase of PittPatt, a face scanning software developed by Carnegie Mellon University’s Robotics Institute in Pittsburgh.  

Though it’s not clear how Google plans to use PittPatt’s technology, which uses algorithms to identify and track human faces in photos and video, PittPatt does have the potential to automatically link a person’s image to their social media profiles, Google+ account and a host of Google products, including Image Search, YouTube and Picasa.

 “From a user’s perspective, automated systems make life easier,” Webb says. “Enabling users to scan objects and people using our mobile phones to gain more information, especially when that technology works, is wonderfully helpful.”

PittPatt works in about 60 seconds, scanning comparative databases to identify a person from publically available photos or videos. It even works if their face is partially covered.  

More startling is the fact that this technology has been demonstrated to accurately predict a person’s Social Security number using profile information such as date of birth. While the threat of identity exploitation is very real for American users, there is no evidence Social Insurance Numbers are at stake for disclosure for Canadians using this technology.

“The (PittPatt) team proved that it’s possible to predict what someone’s social security number is based on their Facebook profile. So if we know that, via Facebook, we can determine who someone is based only on their photo, and we can determine social security numbers based on Facebook profiles, it is likely (and proven) that we can determine someone’s social security number by scanning their face,” Webb explains.

Viewdle’s Social Camera is another new recognition app that uses advanced technology to auto-detect and tag photos. Currently in Beta, the app for Andriod phones creates ‘faceprints,’ tagged data unique to an individual’s facial characteristics. The faceprints can be synced to any device and shared through Facebook, Flickr, MMS or email. Social Camera also automatically matches a tagged photo to that person’s contact information already stored in the user’s phone.

Cause for concern 

As more identification apps enter the market, the challenge will be precision detection, Webb suggests.

“While the (facial recognition) technology is incredibly sophisticated, the failure rate is still high,” she says, adding a warning:  “If you value your privacy, yes, you should be concerned about face and iris recognition technology.”

“We are arguably the most monitored society in the history of civilization,” Webb says of our constant surveillance. “It has become easier to not just see that we’re in a place, but also who we are, where we work and who we’re connected to in real life,” she says.

“It’s good practice to monitor yourself and to ensure that what digital identity information exists about you is information you’re comfortable sharing with the entire world.”

There are other steps users can take to avoid complete identification exploitation.

Consider raising privacy settings to the highest possible security on Facebook, Twitter, MySpace and other social media platforms. Change image sharing settings to exclude others from tagging photos of you.

Avoid publishing personal information such as your address and phone number on social media profiles.

Monitor any developer updates to your social media profiles that may re-set your privacy controls to the lowest or default settings.  

Facial recognition software such as Social Camera and PittPatt will not work to identify you if your social media avatar is not an image of yourself. That is, if your Facebook profile photo is a photograph of your dog, a user who scans your likeness will not be able to recognize you or link instantly to your social media profiles.

CV Dazzle is a prototype application that allows users to edit their image using hair or makeup modifications.

The augmented reality app essentially camouflages a user’s identity, thereby thwarting detection from facial recognition software.

For a list of other technology trends to look forward to in the coming year, watch AmyWebb’s Tech Trends presentation at the 2011 Online News Association conference and online journalism awards in Boston.